summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/importers/crossref.py
diff options
context:
space:
mode:
Diffstat (limited to 'python/fatcat_tools/importers/crossref.py')
-rw-r--r--python/fatcat_tools/importers/crossref.py15
1 files changed, 12 insertions, 3 deletions
diff --git a/python/fatcat_tools/importers/crossref.py b/python/fatcat_tools/importers/crossref.py
index faee6aac..d8abf3eb 100644
--- a/python/fatcat_tools/importers/crossref.py
+++ b/python/fatcat_tools/importers/crossref.py
@@ -11,12 +11,14 @@ from .common import EntityImporter, clean
# The docs/guide should be the cannonical home for these mappings; update there
# first
+# Can get a list of Crossref types (with counts) via API:
+# https://api.crossref.org/works?rows=0&facet=type-name:*
CROSSREF_TYPE_MAP = {
'book': 'book',
'book-chapter': 'chapter',
'book-part': 'chapter',
'book-section': 'chapter',
- 'component': None,
+ 'component': 'component',
'dataset': 'dataset',
'dissertation': 'thesis',
'edited-book': 'book',
@@ -158,6 +160,7 @@ class CrossrefImporter(EntityImporter):
def want(self, obj):
if not obj.get('title'):
+ self.counts['skip-blank-title'] += 1
return False
# do most of these checks in-line below
@@ -174,10 +177,12 @@ class CrossrefImporter(EntityImporter):
if obj.get('type') in (None, 'journal', 'proceedings',
'standard-series', 'report-series', 'book-series', 'book-set',
'book-track', 'proceedings-series'):
+ self.counts['skip-release-type'] += 1
return None
# Do require the 'title' keys to exsit, as release entities do
if (not 'title' in obj) or (not obj['title']):
+ self.counts['skip-blank-title'] += 1
return None
release_type = self.map_release_type(obj['type'])
@@ -376,10 +381,13 @@ class CrossrefImporter(EntityImporter):
# filter out unreasonably huge releases
if len(abstracts) > 100:
+ self.counts['skip-huge-abstracts'] += 1
return None
- if len(refs) > 2000:
+ if len(contribs) > 2000:
+ self.counts['skip-huge-contribs'] += 1
return None
if len(refs) > 5000:
+ self.counts['skip-huge-refs'] += 1
return None
# release date parsing is amazingly complex
@@ -406,6 +414,7 @@ class CrossrefImporter(EntityImporter):
title = clean(obj.get('title')[0], force_xml=True)
if not title or len(title) <= 1:
# title can't be just a single character
+ self.counts['skip-blank-title'] += 1
return None
subtitle = None
@@ -413,7 +422,7 @@ class CrossrefImporter(EntityImporter):
subtitle = clean(obj.get('subtitle')[0], force_xml=True)
if not subtitle or len(subtitle) <= 1:
# subtitle can't be just a single character
- return None
+ subtitle = None
if extra_crossref:
extra['crossref'] = extra_crossref